Retinopathy of prematurity (ROP) is a potentially preventable disorder of the retina that occurs predominantly in preterm newborns. Recent evidence suggests that exposure to systemic inflammation might contribute to the etiology and pathogenesis of ROP. In this application, we propose to (objective #1) measure in existing blood samples from a multi-center, NIH-funded cohort study of extremely low gestational age newborns (ELGAN study, U01 NS40069) 18 protein biomarkers in 1,206 ELGANs and add this information to an existing database of 25 biomarkers we have already measured in 924 of the same 1,206 ELGANs. We will also measure these previous 25 markers in the remainder of 282 newborns. The result will be the currently largest biomarker database worldwide of 43 biomarkers (from postnatal days 1, 7, 14, 21, and 28) available for 1,206 newborns with a gestational age <28wks. The new biomarker measurements will be done by the same ELGAN-Co-investigators who measured the initial 25 biomarkers using the Meso Scale Discovery (MSD) multiplex platform. The additional biomarkers were selected specifically for ROP association studies, based on published evidence that the availability of oxygen, the presence of inflammation, the concepts of angiogenesis and erythropoiesis, as well as pre-existing maternal pregnancy conditions such as preeclampsia all appear to play roles in ROP etiology and pathogenesis. We also propose (objective #2) to subsequently perform epidemiologic (risk) analyses using advanced multivariable biostatistical techniques. We will test two specific null-hypotheses: (1) Biomarkers do not predict severe ROP occurrence; and (2) biomarkers do not predict ROP progression. Both hypotheses will be tested by members of the original ELGAN data analysis team in comprehensive multivariable regression models that allow for appropriate control for confounding and for stratified analyses targeted at the discovery of effect modifiers. Pre hoc power analyses suggest that our sample size is large enough to detect appreciable relative risks with sufficient statistical power. The results from our project will have a significnt impact on our current understanding of ROP pathogenesis and will help pave the way towards new prevention and treatment strategies.